Clarias and Tilapia Interaction in Polyculture

C. Kwei LinAgricultural and Food Engineering Division, Asian Institute of Technology,

G.P.O. Box 2754Bangkok, Thailand

Abstract

Experiments were conducted to assess the effects of polyculture on tilapiagrowth and to evaluate African catfish (Clarias gariepinus) as a predator controlfor unwanted tilapia offspring in a polyculture system. Three treatments weretested using nine 200 m2 earthen ponds. Ponds were stocked with male tilapia intreatment 1 (Tl), mixed sex tilapia in treatment 2 (T2), and mixed sex tilapia andcatfish in treatment 3 (T3). Fish were fed a commercial diet ration of 1.5% of bodyweight per day twice a day at 0900 hr and 1500 hr. Feed inputs, adjustedbiweekly, were based on Nile tilapia (Oreochromis niloticus) biomass. Gutcontents of African catfish were analyzed two weeks prior to termination of theexperiment. Water quality parameters were estimated biweekly. Diel analyses ofdissolved oxygen (DO), temperature, and pH were performed at three depthsand six time intervals, and net primary productivity (NPP) was estimated usingDO concentrations. A t-test was used to compare final fry biomass, daily weightgain (DWG) of Nile tilapia, net fish yield (NFY), and the diel variability of gutcontent. DWG and NFY were significantly higher for treatment T1 in which maleNile tilapia were stocked. Mean fry biomass for each treatment showed largefluctuations. Statistical analyses of two of the three replicates of T1, T2, and allthree replicates of T3 indicated that T2 production was significantly greater thanT3 production. Forty-seven percent of the sampled African catfish contained oneof the following in the stomach or intestines: whole fish, partially digested fish,fish flesh, fish scales, fins, or bones; twenty-six percent of the African catfish hadfreshly ingested fry in the stomach. Biweekly variation in mean values for waterquality parameters were comparable for each treatment. Tilapia production inpolyculture with African catfish was significantly lower than the culture systemswith either all male tilapia or mixed sex tilapia. African catfish predation oftilapia fry was sufficiently effective to serve as a population control for tilapia;however, active predation occurred only in semi-intensive culture systems wherefish were fed their natural diet.

Introduction

Overcrowding in pond culture caused by the prolific reproduction of tilapia(Oreochromis niloticus) results in competition for food and consequent yieldscomposed mainly of small fish of low market value. One of the managementstrategies used to control pond overpopulation is tilapia-snake head (Channastriatus) polyculture (Kaewpaitoon, 1992). The carnivorous snake head not onlyfunctions as a biological control, but also increases economic gain due to its highmarket value.

African catfish (Clarias gariepinus), which often contaminate tilapia pondsin Egypt, have been recently investigated as a potential aquaculture species.Considerable effort has been made to investigate its reproductive biology (DeGroot, 1987). Feeding habits were also described. It is omnivorous, feedingmainly on detritus, invertebrates, and small fish; the extent of its predatoryability is unknown (Pillay, 1990). This experiment was designed to: 1) determinethe suitability of African catfish as a predator to control unwanted tilapiaoffspring in a polyculture system, and 2) to determine the effect of polycultureon tilapia growth.

Materials and Methods

Pond Preparation and StockingThe experiment was to be conducted at the Central Laboratory for

Aquaculture Research at Abbassa, Egypt. Due to a shortage of catfish fry inEgypt and logistical problems, the experiment was moved to the Asian Instituteof Technology (AIT) in Thailand, another PD/A CRSP site. This site, locatedapproximately 40 km north of Bangkok, has a tropical climate.

The experiment was conducted from 4 June 1994 to 5 October 1994. Nine200 m2 earthen ponds were used for the experiment. Three replicates wereperformed for the following treatments:

Table 1. Stocking treatments of mixed-sex Nile tilapia and African catfish.

Treatment 1 (T1): ponds stocked with male tilapia onlyTreatment 2 (T2): ponds stocked with mixed sex tilapiaTreatment 3 (T3): ponds stocked with mixed sex tilapia and catfish.

Prior to commencing the experiment, ponds were drained, sun-dried, andfilled up to a depth of 80-90 cm with water from the peripheral canal. Waterlevels were maintained weekly to compensate for evaporation and seepagelosses. Each pond was fertilized with industrial grade urea (45% N) and triple

super phosphate (21% P) at 21 g and 7 kg/ha/week, respectively. Fertilizerswere dissolved in water and then spread over the pond surface.

Juvenile Nile tilapia were obtained from the AIT hatchery. One week priorto stocking, the sex was visually determined and male and female fish weremanually sorted into separate hapas. Due to large variations in individualweight, male tilapia were manually sorted into two size groups. Each treatmentreceived large and small fish at a ratio of 2: 1, respectively. Tilapia were stockedat a density of 2 fish/m2 (400 fish/pond) in all three treatments. The sex ratios oftilapia in treatments T2 and T3 were maintained at approximately a 1:1 ratio.

Swim-up African catfish fry were obtained from a local commercialhatchery and reared in AIT ponds until they reached at least 50 g. During thisrearing period, the catfish fry were fed with small floating pellets to satiationtwice per day at 0900 hr and l500 hr. A larger pellet was used after the fish hadgrown. The proximate nutrient composition of both pellet sizes was thesameÑcrude protein 30%, crude lipids 4%, ash 8%, and moisture 12%.

The following measurements were taken during stocking of each pond:individual lengths and weights of 15 large male tilapia (T1), 10 small male tilapia(T1, T2, and T3), 20 female tilapia (T2 and T3), 15 African catfish (T3), and totalweight of the fish remaining to be stocked. African catfish were introduced intothe culture system two weeks after stocking of the tilapia to ensure that Niletilapia fry were abundant in the culture system at the time of the African catfishintroduction. Details of the stocking for each treatment are presented in Table 1.

During the first six weeks of the study fish were fed a commercial dietration of 1.5% of body weight per day (Table 2). Equal rations were offered twicea day at 0900 hr and 1500 hr. Feed input was adjusted biweekly based on Niletilapia biomass. To monitor growth, Nile tilapia and African catfish weresampled for individual length and weight measurementsÑ40 male tilapia (T1),20 male and 20 female Nile tilapia (T2), and 20 male and 20 female Nile tilapia,and 15 African catfish (T3). Feeding was discontinued after six weeks because thecatfish consumed feed instead of tilapia fry.

Gut Content Analysis of African CatfishAbout two weeks prior to termination of the experiment, catfish were

sampled for a gut contents analysis to determine if tilapia fry were an importantcomponent of catfish diet. Ponds were seined four times a day at 0600 hr. 1300hr. 1600 hr. and 2300 hr to investigate gut content fluctuations over time. Twocatfish were selected from each pond at each time interval. Fish were killedimmediately by exposing them to a high concentration of Benzocaine (about 300ppm). Sampling was repeated after ten days.

After total length and weight were determined, the fish were dissected.The alimentary canal was removed without rupture or content loss and weighed.Gut contents were preserved in 4% formalin. The weight of gut content wasdetermined as the difference between the weight of the full alimentary canal andthe weight of the empty alimentary canal. Gut content was expressed aspercentage of total body weight [(content weight/total body weight) X 100]. Thegut content was visually inspected for fry residue and microscopically inspectedfor fish scales and/or bones. Other components such as plant materials andplankton were also recorded if they were the dominant component of the gut

contents.

Table 2. Commercial ration feed amounts applied during the first six weeks ofthe study.

Water Quality AnalysisStandard methods (APHA, 1985) were used to estimate water quality

parameters biweekly. One integrated water column sample was collected fromeach pond with a column water sampler and analyzed for total ammonia, nitrite,and nitrate, total Kjeldahl nitrogen, orthophosphate, total phosphorous,alkalinity, total suspended solids (TSS), and total volatile solids (TVS). Watertemperature and dissolved oxygen (DO) were measured in situ with a digitalmeter (YSI model 57). Pond water pH was determined using a Suntex digitalmeter model SP-5A.

Diel analysis of DO, temperature, and pH was carried out at three depthsat six time intervals (0600 hr, 0900 hr, 1200 hr, 1600 hr, 1800 hr, 2300 hr, and 0600hr the following day) each month. Net primary productivity (NPP) wasestimated using DO concentrations between 1600 hr and the preceding 0600 hr.

Statistical AnalysisTotal crop weight was measured at the time of stocking and again at

harvest. A t-test (STATGRAPHICS version 5) was used to compare final frybiomass, daily weight gain (DWG) of Nile tilapia, net fish yield (NFY), and thediel variability of gut content. Further investigated was the DWG of Nile tilapiaand African catfish during feeding and after feeding was terminated (a = 0.05).Collection schedule details for fish sampling, water quality analysis, pondfertilization, water level maintenance, and feed amount (kg/pond/day) for thefirst six weeks are presented in Table 3.

Results

Fish Survival, Growth, Production, and Fry BiomassNile tilapia mean population weight in all treatments ranged from 28.3-

32.9 g and African catfish mean population weight in T3 ranged from 67.7-75.1 g.

Table 3. Collection schedule details for fish stocking, sampling, harvest, waterquality analysis, and pond fertilization.

Initial and final mean weight, daily weight gain (DWG), survival, net fish yield(NFY), and fry biomass of Nile tilapia are presented in Table 4. Fortnightly meanweight increases of Nile tilapia and African catfish are presented in Figures 1 and2, respectively. Mean NFY of Nile tilapia for each treatment are presented inFigure 3.

DWG and NFY of the Nile tilapia populations were significantly higherfor treatment T1 (stocked with male Nile tilapia) than treatments T2 and T3(p<0.05). However, there were no significant differences between Nile tilapiapopulations for DWG and NFY in T2 and T3. The well established fact that maleNile tilapia in a monosex population grow faster than individuals in a mixed-sexpopulation may account for this finding.

The mean fry biomass for each treatment is presented in Figure 4. Largefluctuations among replicates of the treatments were observed. The t-testindicated that there were no significant differences in mean fry biomass fordifferent treatments when all three replicates for each treatment were taken intoaccount. It is important, however, to note that one of three replicates fortreatments T1 and T2 showed much lower Nile tilapia fry biomass comparedwith the two remaining replicates. This may have been due to the unintentionalpresence of snake head that consumed fry in the ponds of the culture system.There were seven snakeheads in treatment T2 with a mean weight of 292.3 ± 37.4g (SE) in one of the ponds (P#12), and three snakeheads in treatment T1 with amean weight of 429.6 ±145.9 g (SE) in another pond (P#5). Snakeheads are wellknown predators and most likely reduced the fry abundance in the replicates ofT1 and T2, which had lower Nile tilapia fry biomass than treatments T1 and T2.Statistical analyses were therefore made using only two replicates fromtreatments T1 and T2 and three replicates from treatment T3. The fry productionin treatment T2 was significantly greater than fry production in treatment T3indicating that active catfish preyed on tilapia fry.

Gut Analysis of African CatfishGut content of African catfish, sampled four times per day, was calculated

as a percent of the body weight of the fish (see Figure 5 and Table 5). Gut contentgenerally decreased from 0600 hr to 1800 hr. but increased from 1800 hr to 2300hr. Gut content at 2300 hr. however, generally was lower when compared withthe preceding 0600 hr sample. This observation may be related to the nocturnalnature of the feeding cycle of African catfish. Greater feeding activity during thenight and digestion during the day causes the reduction of the gut content aftersunrise. The t-test indicated significantly lower gut content at 1800 hr comparedwith 0600 hr. No significant differences were found for gut contents at othertimes of the day.

Gut content analysis also revealed that, on average, 46.7 % (± 8.2% SD) ofthe sampled African catfish had one of the following in the intestine or stomach:whole fish, partially digested fish, fish flesh, fish scales, fins, or bones. Only26.4% (± 6.6% SD) of the African catfish, however, had freshly ingested fry in thestomach. The percentage of captured African catfish having Nile tilapia fry in thestomach at various times of the day is shown in Figure 6 and Table 5. There wasno significant difference in the percentage of African catfish having Nile tilapia

fry in the gut at different times of the day.

Water Quality During the ExperimentBiweekly variation in mean values (± 1 SE) for the water quality

parametersÑdissolved oxygen, pH, alkalinity, total Kjeldahl nitrogen, ammonianitrogen, total phosphorous and total suspended solidsÑis presented for eachtreatment throughout the experimental period (Figures 7a-g). Overall,comparable values for different treatments were observed. Ranges of themaximum and minimum values for different water quality parameters ofdifferent treatments are summarized in Table 6.

Table 4. Initial and final mean weights, daily weight gain, survival, net fish yield,

and fry biomass of Nile tilapia (± 1 SE) for each treatment.

Fig.1. Fortnightly increase in Nile tilapia mean weight (± 1 SE) for eachtreatment.

Fig. 2. Fortnightly increase in African catfish mean weight (± 1 SE) for eachtreatment.

Fig. 3. Mean net fish yield (± 1 S.E.) for each treatment.

Fig. 4. Net fish yield of Nile tilapia (± l SE) for each treatment.

Fig. 5. Gut content expressed as % of body weight of African catfish at differenttimes of the day.

Fig. 6. Percentage of sampled African catfish having Nile tilapia fry in stomach.

Table 5. Details of gut content analysis of African catfish.

Table 6. Range values of water quality parameters for each treatment during

culture period.

Fig. 7a. Biweekly variation in mean value for the dissolved oxygen water qualityparameter.

Fig. 7b. Biweekly variation in mean value for the pH water quality parameter.

Fig. 7c. Biweekly variation in mean value for the alkalinity water qualityparameter

Fig. 7d. Biweekly variation in mean value for the total kjeldhal nitrogen waterquality parameter.

Fig. 7e. Biweekly variation in mean value for the ammonia nitrogen water qualityparameter.

Fig. 7f. Biweekly variation in mean value for the total phosphorous water quality

parameter.

Fig. 7g. Biweekly variation in mean value for the total suspended solids waterquality parameter.

Conclusions

1. The reliability of manually sexing tilapia is low and results in seriouscontamination of all male treatments with female fish. Reproductionundermined the all male treatment experiment.

2. Production of large adult tilapia is significantly lower in ponds with mixedsexes; however, the total biomass including reproduction does not differgready.

3. Tilapia production in polyculture with African catfish is significantly lowerthan the culture systems with either all male tilapia or mixed sex tilapia.

fi

4. African catfish predation of tilapia fry is sufficiently effective to serve aspopulation control for tilapia. The active predation, however, occurred onlyin semi-intensive culture systems where fish are fed their natural diet.

5. Gut content analysis of catfish can provide a reliable method for estimatingpredation.

6. Further research should be focused on the following areas: 1) the assessmentof the effects of increased variation in tilapia and catfish stocking ratios, and2) the bioeconomics of the polyculture versus the monoculture of catfish andtilapia.

Literature Cited

APHA (American Public Health Association), 1989. American Water WorksAssociation and Water Pollution Control Federation: Standard methods forthe examination of water and wastewater, 17th edition. APHA, Washington,D. C.

De Groot, S.J., 1987. Culture of Clarias species. Elsevier Science Publishers,Amsterdam, 366 pp.

Kaewpaitoon, K., 1992. Utilization of septage-raised tilapia (Oreochromis niloticus)as feed for snakehead (Channa striates). PhD Dissertation. Asian Institute ofTechnology, Bangkok, Thailand, 212 pp.

Pillay, T.V.R., 1990. Aquaculture: Principles and Practices. Fishing News Books,London, 575 pp.